Re: Projections and Media incentives
From CNN.com: Sen. John Kerry's campaign refused to concede early Wednesday, arguing that more than 250,000 provisional and absentee ballots remained to be counted in the key battleground state of Ohio. New Mexico will not release presidential election results until later Wednesday because thousands of absentee ballots remain uncounted, according to a spokesman for the secretary of state. The vote also remained unsettled in Iowa, where broken machines, a delay in opening absentee ballots and a delay in the arrival of other absentee ballots were slowing completion of the secretary of state's final count in the presidential election, elections officials said early Wednesday. So there are a large number of ballots remaining. However, in Ohio, it is hard to see how Kerry gets over 2/3 of the remaining votes. -Jeff David Friedman [EMAIL PROTECTED] 11/03/04 02:36AM At the moment, according to CNN, Bush is leading Kerry in New Mexico by about 28,000 votes out of a little over 500,000. 97% of the precincts have reported. Assuming, as I believe is usually the case, that precincts are all about the same size, that means that there are only about 15,000 votes uncounted. If Kerry gets every one of them, he still loses. But CNN (and, last time I checked, CBS) show New Mexico as still undecided. The situation in Ohio is similar. Bush is ahead by about 140,000 votes out of a little over five million, with 98% of precincts reporting. That means that there are about a hundred thousand votes still to be counted. If Kerry gets all of them, he still loses. Ohio is shown as too close to call. It's possible that I'm missing something--that there is some additional large source of uncertainty that could change the outcome, such as huge numbers of uncounted absentee ballots--but I don't think so. It looks to me as though CNN doesn't want to concede that the election is over because when they do people will stop viewing their web site (and, I presume, watching their TV programs). A secondary element is that, if they call a state for one candidate and it goes for another, they look bad. If they just delay, they have never been wrong in the same clear sense. I suspect that is the reason, earlier, that CNN failed to call Florida for Bush until hours after, on their figures, it was obvious that he was going to carry it. I can understand that the media are a bit reluctant to commit themselves on that particular state. -- David Friedman www.daviddfriedman.com
Re: Projections and Media incentives
[EMAIL PROTECTED] (David Friedman) writes: It's possible that I'm missing something--that there is some additional large source of uncertainty that could change the outcome, such as huge numbers of uncounted absentee ballots--but I don't think so. It looks to me as though CNN doesn't want to concede that the election is over because when they do people will stop viewing their web site (and, I presume, watching their TV programs). That strategy would normally fail due to viewers switching to the news source that is first to report the result. Maybe Kerry voters want to avoid knowing the news, or CNN mistakenly thinks that is the case? Kerry has apparently conceded just now, while CNN is only calling 254 electoral votes for Bush, and Fox showing 269 for Bush (strangely reluctant to call Nevada). -- -- Peter McCluskey | To say that President George W. Bush has been www.bayesianinvestor.com | spending money like a drunken sailor is an insult to | drunken sailors. - Nick Gillespie, Reason Magazine
Re: Incentives
Psychologists have conducted experiments where the subjects are (randomly) split into two categories. They both perform the same task, perhaps a memory drill, and then one group gets paid money for participating and the other doesn't. After the experiment, i.e. the task that the subjects were told was the experiment, the subjects are interviewed. One of the questions asks how much they enjoyed the experiment. Subjects who were paid money enjoy the task significantly less than those who aren't. The theory behind this is that when a person does the task, their mind needs a reason to avoid cognitive dissonance. When they are paid, the money acts as the reason; when they aren't paid, enjoying the task acts as the reason. To put another way, one's mind imposes enjoyment ex post, so that it doesn't have to cope with the disconnect of doing something for no good reason and disliking doing it. Hazing rituals are supposed to perform a similar function. If one puts up with the hazing, it must be for a good reason. Therefore, the group that does the hazing, the frat, military academy, or whatever, is seen in a better light to avoid the cognitive dissonance. Don't judge this theory based on my explanation of it. As I've noted before, I'm a clumsy writer at best. But that is the theory as I recall it. Whether grades fit the theory, I haven't a clue. Hope that helps, -jsh --- [EMAIL PROTECTED] wrote: The following appeared in an article on grade inflation in the Chronicle of Higher Education: Grades motivate (a fallacy according to the article). With the exception of orthodox behaviorists, psychologists have come to realize that people can exhibit qualitatively different kinds of motivation: intrinsic, in which the task itself is seen as valuable, and extrinsic, in which the task is just a means to the end of gaining a reward or escaping a punishment. The two are not only distinct but often inversely related. Scores of studies have demonstrated, for example, that the more people are rewarded, the more they come to lose interest in whatever had to be done in order to get the reward. (That conclusion is essentially reaffirmed by the latest major meta-analysis on the topic: a review of 128 studies, published in 1999 by Edward L. Deci, Richard Koestner, and Richard Ryan.) Is anyone on the list familiar with this literature? It sounds like they are saying that incentives don't matter. Cyril Morong __ Do you Yahoo!? Yahoo! Web Hosting - Let the expert host your site http://webhosting.yahoo.com
Re: Incentives
I enjoyed Alfie Kohn's book Punished by Rewards, which is a popularization of much of this research. http://www.amazon.com/exec/obidos/tg/detail/-/0618001816/qid=1037396034/sr=8-1/ref=sr_8_1/104-2976940-3732712?v=glances=booksn=507846 I haven't read them yet, but Deci and Dweck seem to be a couple of the principal researches in the effects of external rewards on intrinsic motivation. Why We Do What We Do: Understanding Self-Motivation Edward L. Deci, Richard Flaste http://www.amazon.com/exec/obidos/tg/detail/-/0140255265/ref=pd_bxgy_text_1/104-2976940-3732712?v=glances=books Self-theories: Their Role in Motivation, Personality, and Development (Essays in Social Psychology) Carol S. Dweck http://www.amazon.com/exec/obidos/tg/detail/-/1841690244/qid=/sr=/ref=cm_lm_asin/104-2976940-3732712?v=glance Chris john hull wrote: Psychologists have conducted experiments where the subjects are (randomly) split into two categories. They both perform the same task, perhaps a memory drill, and then one group gets paid money for participating and the other doesn't. After the experiment, i.e. the task that the subjects were told was the experiment, the subjects are interviewed. One of the questions asks how much they enjoyed the experiment. Subjects who were paid money enjoy the task significantly less than those who aren't. The theory behind this is that when a person does the task, their mind needs a reason to avoid cognitive dissonance. When they are paid, the money acts as the reason; when they aren't paid, enjoying the task acts as the reason. To put another way, one's mind imposes enjoyment ex post, so that it doesn't have to cope with the disconnect of doing something for no good reason and disliking doing it. Hazing rituals are supposed to perform a similar function. If one puts up with the hazing, it must be for a good reason. Therefore, the group that does the hazing, the frat, military academy, or whatever, is seen in a better light to avoid the cognitive dissonance. Don't judge this theory based on my explanation of it. As I've noted before, I'm a clumsy writer at best. But that is the theory as I recall it. Whether grades fit the theory, I haven't a clue. Hope that helps, -jsh --- [EMAIL PROTECTED] wrote: The following appeared in an article on grade inflation in the Chronicle of Higher Education: Grades motivate (a fallacy according to the article). With the exception of orthodox behaviorists, psychologists have come to realize that people can exhibit qualitatively different kinds of motivation: intrinsic, in which the task itself is seen as valuable, and extrinsic, in which the task is just a means to the end of gaining a reward or escaping a punishment. The two are not only distinct but often inversely related. Scores of studies have demonstrated, for example, that the more people are rewarded, the more they come to lose interest in whatever had to be done in order to get the reward. (That conclusion is essentially reaffirmed by the latest major meta-analysis on the topic: a review of 128 studies, published in 1999 by Edward L. Deci, Richard Koestner, and Richard Ryan.) Is anyone on the list familiar with this literature? It sounds like they are saying that incentives don't matter. Cyril Morong __ Do you Yahoo!? Yahoo! Web Hosting - Let the expert host your site http://webhosting.yahoo.com
Incentives
The following appeared in an article on grade inflation in the Chronicle of Higher Education: "Grades motivate (a fallacy according to the article). With the exception of orthodox behaviorists, psychologists have come to realize that people can exhibit qualitatively different kinds of motivation: intrinsic, in which the task itself is seen as valuable, and extrinsic, in which the task is just a means to the end of gaining a reward or escaping a punishment. The two are not only distinct but often inversely related. Scores of studies have demonstrated, for example, that the more people are rewarded, the more they come to lose interest in whatever had to be done in order to get the reward. (That conclusion is essentially reaffirmed by the latest major meta-analysis on the topic: a review of 128 studies, published in 1999 by Edward L. Deci, Richard Koestner, and Richard Ryan.)" Is anyone on the list familiar with this literature? It sounds like they are saying that incentives don't matter. Cyril Morong
Re: Micro Incentives -- real case question
Dear Armchairists (?), My MA advisory company has a few sector teams (in Vienna London). We also have a network of local offices in all the CEE capital cities. On a big Bank deal, for instance, there will be a local team element, and a London sector team element. When we get the mandate, it's partly because our expensive London expert team is recognized as fine, but also because our local team (in Bratislava, for instance), is also strong. Internally we split the fees between the two teams -- which is annually important for bonuses. We are not satisfied with the internal bickering about fee splits and work -- on any given mandate, both teams (local expert) believe they could basically do it all alone. My top management would like an incentive/accounting system which satisfies both teams, and gets both teams to cooperate more fully, and to use the advantages of each more fully. Is there any micro- / game-theory analysis on such internal business organization? Recommendations suggestions welcomed. Tom Grey CA IB Securities One interesting possibility is to give each team all the money. This assumes that the benefit to the firm from the deal is measurable. Consider a simple case: Two individuals produce a product; it's total value is a function of inputs by each. Inputs are not measurable, but total value is. How do we give each the right incentives? Say that total value varies, on average, from 0 to $100, average $50. Each individual puts $25 into a kitty in advance. Each individual then gets back the full value of the output. Obviously the kitty has to be held by something like an insurance company, since it is breaking even only on average. In your context, this presumably works out as lowering the salaries paid to your people and increasing the bonus so that total bonus paid out for a transaction is twice the actual value. Of course, it only works if their salary is in part paying for something other than such transactions--otherwise they would have to pay the firm for the privilege of working there (and getting a shot at the bonuses). The underlying logic of this approach depends on two facts: 1. Each party makes the optimal investment if, on the margin, he gets the full benefit from his investment. 2. Marginal doesn't have to equal average. You can find a different version of the same argument (with negative payoffs due to auto accidents and the like) in the chapter on torts, I think pp. 195 and 203-4. -- David Friedman Professor of Law Santa Clara University [EMAIL PROTECTED] http://www.daviddfriedman.com/
Re: Micro Incentives -- real case question
My top management would like an incentive/accounting system which satisfies both teams, and gets both teams to cooperate more fully, and to use the advantages of each more fully. Is there any micro- / game-theory analysis on such internal business organization? Lots! But you would have to be more specific about the nature of the product, how output is evaluated, how contracts are secured, what the nature of cooperation between the two groups is etc. before anyone would know how to apply it. Also, some people get paid for doing this sort of thing and you might keep in mind the old line you get what you pay for. - - Bill Dickens William T. Dickens The Brookings Institution 1775 Massachusetts Avenue, NW Washington, DC 20036 Phone: (202) 797-6113 FAX: (202) 797-6181 E-MAIL: [EMAIL PROTECTED] AOL IM: wtdickens